Socioeconomic inequalities in child welfare interventions are five times stronger in unequal and affluent local authorities than they are in equal and deprived local authorities. Why?

Social Work sociology

One of the articles we had published in Children and Youth Services Review this year highlighted a new finding that the social gradient in child welfare interventions was predicted to be as much as five times stronger in local authorities with both lower-than-average deprivation and higher-than-average income inequality than it was in local authorities with higher-than-average deprivation and lower-than-average income inequality. This blog post explores one reason why this might be.

Calum Webb
2020-08-20


One of the articles we had published in Children and Youth Services Review this year highlighted a new finding that the social gradient in child welfare interventions was predicted to be as much as five times stronger in local authorities with both lower-than-average deprivation and higher-than-average income inequality than it was in local authorities with higher-than-average deprivation and lower-than-average income inequality.

This blog post explores one reason why this might be, but before discussing this, it’s important to review what we mean by the social gradient, income inequality, how we measured these, and how local income inequality differs from many of the ways we think about ‘inequality’.

The technical bit…

The social gradient in child welfare interventions is a measure of the how much intervention rates - such as children being placed on child protection registers, or being taken into care - differ according to family socioeconomic status, which is largely measured by their economic and material wellbeing. Sometimes this is measured at different levels so, for example and in our ‘Untangling Child Welfare Inequalities’ paper, we look at the social gradient in neighbourhood (approx 1500 people) intervention rates based on neighbourhood deprivation levels.

In order to do this, and to standardise certain things to control for ‘confounding factors’ - that is, other features of neighbourhoods the effect of which we want to control for so we don’t confuse it with the deprivation effect - we created a way to express the social gradient in a single number that represents how many times we would expect the rate of child welfare interventions to increase or decrease for every one standard deviation increase in deprivation, measured by the Indices of Multiple Deprivation. A one-standard deviation difference is about the same to comparing an ‘average’ deprivation neighbourhood with one that is ranked in the middle of the third most deprived of all neighbourhoods.

For simplicity’s sake, I refer to a minus 1 change in standard deviation as an ‘average low deprivation’ neighbourhood, and plus one standard deviation change as representing an ‘average high deprivation’ neighbourhood. It’s best to not overthink it too much and just consider it us comparing low, average, and high deprivation neighbourhoods in a way that doesn’t mean we focus on the extremes. See the graph below for a more simple example of local authority deprivation. We scaled up the same logic to local authorities - so an ‘average high deprivation local authority’ would be one with a deprivation score one standard deviation, and did the same for ‘income inequality’.